def test_config(self):
num_heads = 12
key_dim = 64
test_layer = attention.KernelAttention(
num_heads=num_heads,
key_dim=key_dim,
feature_transform='exp',
num_random_features=128,
is_short_seq=True)
new_layer = attention.KernelAttention.from_config(
test_layer.get_config())
# If the serialization was successful, the new config should match the old.
self.assertAllEqual(test_layer.get_config(), new_layer.get_config())
def test_attention_scale_by_length(self, seq_length):
num_heads = 12
key_dim = 64
batch_size = 2
test_layer = attention.KernelAttention(
num_heads=num_heads,
key_dim=key_dim,
num_random_features=0,
scale_by_length=True)
query = tf.random.normal(
shape=(batch_size, seq_length, key_dim))
value = query
encoder_inputs_mask = tf.ones((batch_size, seq_length), dtype=tf.int32)
masks = tf.cast(encoder_inputs_mask, dtype=tf.float32)
output_scale_by_length = test_layer(
query=query, value=value, attention_mask=masks)
test_layer._scale_by_length = False
output_no_scale_by_length = test_layer(
query=query, value=value, attention_mask=masks)
if seq_length == 512: # Equals because log(seq_length, base=512) = 1.0
self.assertAllClose(output_scale_by_length, output_no_scale_by_length)
else:
self.assertNotAllClose(output_scale_by_length, output_no_scale_by_length)
def test_attention_projection(self, feature_transform, num_random_features,
training, redraw, is_short, begin_kernel):
num_heads = 12
key_dim = 64
seq_length = 1024
batch_size = 2
test_layer = attention.KernelAttention(
num_heads=num_heads,
key_dim=key_dim,
feature_transform=feature_transform,
num_random_features=num_random_features,
redraw=redraw,
is_short_seq=is_short,
begin_kernel=begin_kernel)
query = tf.random.normal(shape=(batch_size, seq_length, key_dim))
value = query
encoder_inputs_mask = tf.zeros((batch_size, seq_length),
dtype=tf.int32)
masks = tf.cast(encoder_inputs_mask, dtype=tf.float32)
output = test_layer(query=query,
value=value,
attention_mask=masks,
training=training)
self.assertEqual(output.shape, [batch_size, seq_length, key_dim])
def test_redraw_true_no_projection(self):
with self.assertRaisesRegex(
ValueError, "There is nothing to redraw when num_random_features.*"):
_ = attention.KernelAttention(
num_heads=2, key_dim=64, feature_transform="elu",
num_random_features=0, redraw=True)
def test_unsupported_feature_transform(self):
with self.assertRaisesRegex(ValueError, "Unsupported feature_transform.*"):
_ = attention.KernelAttention(feature_transform="test")